SHF: SMALL: Effective and Equitable Technical Interviews in Software Engineering

在软件行业，软件工程师在申请工作时必须通过一系列技术评估，称为技术面试。一种流行的技术面试形式要求软件工程师在面试官面前当场编写计算机代码。业界普遍担心，这种面试可能缺乏现实意义，在评估技术能力方面效果不佳，同时会带来系统性偏见和不必要的压力。该项目旨在开发更公平的软件工程能力的技术评估。主要目标是扩大软件工程劳动力的参与。 为了实现这一目标，该研究调查了涉及人类对代码判断的软件工程过程，例如在代码审查期间，该项目将开发技术来衡量编程任务的技术性能，并确定对编写代码的能力产生负面影响的因素，例如高认知负荷，情感状态差，以及有限的启示。本项目将研究三个主要研究任务。第一项任务将涉及数据挖掘技术访谈经验和利益相关者的参与，以确定有关代码技术评估的综合因素清单。第二项任务将涉及一系列随机对照试验，以确定各种条件的影响，例如解决问题的复杂性、同时大声思考的使用以及刻板印象威胁的存在对技术表现的影响。第三项任务将涉及开发干预措施，以减少对技术性能的不良影响，例如回顾性大声思考或私人访谈，并将包括评估行业试验。总的来说，该项目与寻找机会以提高软件工程和信息工作的公平性的研究流相联系：提高团队的多样性和任期，并为弱势和低资源求职者提供资源。该奖项反映了NSF的法定使命，并通过使用基金会的智力价值和更广泛的影响审查标准进行评估，被认为值得支持。

In the software industry, a software engineer must pass a series of technical evaluations, called technical interviews, when applying for a job. A popular form of the technical interview requires software engineers to write computer code on the spot in front of interviewers. There are industry-wide concerns that such interviews may lack real-world relevance and be ineffective at evaluating technical competencies, while introducing systemic bias and unnecessary stress. This project aims to develop technical evaluations of software-engineering proficiency that are more equitable. The main objective is to enable broadening participation in the software-engineering workforce. To achieve this objective, the research investigates software-engineering processes involving human judgment of code, such as during code reviews.This project will develop techniques to measure the technical performance of a programming task and identify factors that negatively impact the ability to write code, such as high cognitive load, poor affective state, and limited affordances. Three main research tasks will be investigated for this project. The first task will involve data mining technical-interview experiences and engagement with stakeholders in order to determine a comprehensive list of factors concerning technical evaluation of code. The second task will involve a series of randomized, controlled trials to establish the effects of various conditions, such as problem-solving complexity, use of concurrent think-aloud, and presence of stereotype threat on technical performance. The third task will involve the development of interventions that reduce unwanted effects on technical performance, such as retrospective think-aloud or private interviews, and will include evaluating trials in industry. Overall, this project connects to the stream of research on identifying opportunities for increasing equity in software engineering and information work: improving diversity and tenure of teams and providing resources for disadvantaged and low-resource job seekers.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.